In recent years, as data management methods, data management methods using in combination a relational database (RDB) and a key value store (KVS) which is an in-memory have become widespread. The data management methods can shorten the time required for a client terminal to acquire desired data, as compared with a data management method using an RDB only.
In general, in a data management method using an RDB only, it is necessary to search a larger number of tables stored in the RDB for a table including desired data, in response to an access of a client terminal, and acquire the desired data from the table. Inevitably, it takes much time for the client terminal to acquire the desired data.
On the other hand, in a data management method using an RDB and a KVS which is an in-memory, first, it is confirmed whether desired data is stored as a cache in the KVS. In the KVS, i.e., the in-memory, the desired data and a key for acquiring the desired data are stored in a one-to-one correspondence. Thus, if the desired data is stored as a cache, the client terminal can dramatically more speedily acquire the desired data than in the data management method using the RDB only. If the desired data is not stored as a cache, the client terminal will change the storage to be access to the RDB, and acquire the desired data from the RDB. However, since the desired data acquired at this time is registered as a cache in the KVS, from then on, the client terminal can acquire the desired data in a shorter time.
In such a manner, the data management method using the RDB and the KVS in combination enables the client terminal to acquire desired data in a shorter time than the data management method using the RDB only.
However, in the case where the RDB and the KVS are used in combination, if desired data is not stored as a cache in the KVS, the client terminal itself needs to change the storage to be accessed to the RDB. This is inconvenient. To be more specific, a user who operates the client terminal needs to give a description for changing the storage to be accessed from the KVS to the RDB.
Also, in the data management method using the RDB and the KVS in combination, there is a possibility that a disadvantage will be caused when data stored in the RDB is updated. In general, in the case where data stored in the RDB is updated, if data identical to the stored data is stored as a cache in the KVS, it also needs to be updated. However, it is complicated to give a description indicating an operation for causing the cache stored in the KVS to be updated in accordance with the update of the data stored in the RDB. Under present circumstances, generally, the above operation is not performed. Thus, if the data stored in the RDB is updated, but data stored as a cache in the KVS is not updated, there is a possibility that the client terminal will acquire the data not updated.